Architecture and Phyllotaxis of Anisophyllea disticha (Rhizophoraceae)

Architecture and Phyllotaxis of
Atiisophyllea disticha (Rhizophoraceae)

J. R. VINCENT and P. B. TOMLINSON

Harvard Forest, Harvard University, Petersham, U.S.A.

Conspectus

Page

ABSTRACT 3

INTRODUCTION 3

MATERIAL AND METHODS 5

RESULTS 6

Architecture 6

Leaf morphology 8

Leaf anatomy 10

Nodal anatomy 12

Internode length 13

DISCUSSION 14

Phyllotaxy 14

Adaptive considerations 16

BIBLIOGRAPHY 17

Abstract

Anisophyllea disticha does not show distichous phyllotaxis. Erect (orthotropic) shoots with y spiral
phyllotaxis and radial symmetry give rise to tiers of sylleptic branches at regular intervals. The branches
are horizontal (plagiotropic) and have marked dorsiventral symmetry. Their phyllotaxis is unique and
consists of four ranks of alternately arranged leaves, two ranks of scale leaves on the dorsal side and two
ranks of foliage leaves on the ventral side, dorsal and ventral leaves of the two ranks alternating regularly
along the stem on opposite sides. Homology between the three kinds of leafy appendage is indicated by
their constant unilacunar node, but dorsal (adaxial) scales on plagiotropic axes do not subtend axillary
buds. The leaf arrangement is assumed to maximize photosynthesis and corresponds closely to ideal
systems established by theoretical considerations.

Introduction

In the present study attention is drawn to some aspects of morphology and
anatomy of a species of Anisophyllea, A. disticha (Jack)Baillon which allows a bet¬
ter comparison with its putative relatives. The genus Anisophyllea R. Br. ex Sabine
includes about 25 species of trees and shrubs distributed primarily in tropical Africa,
Ceylon, India and Southeast Asia (Ding Hou, 1958). Its relatively recent discovery
in South America (Sandwith, 1952) where 3 species are presently known (Pires and
Rodrigues, 1971) makes its known range almost pan-tropical. Species vary from
treelets of the lower forest storey to tall, canopy trees.

The taxonomic position of Anisophyllea is somewhat controversial. Most authors
have included it in the Rhizophoraceae, usually within a separate tribe

3

Plate 1. Anisophyllea disticha

A. Habit of sapling, obliquely from above.

B. Terminal tier of 5 plagiotropic branches from above, apex of orthotropic axis quiescent.

C. Part of plagiotropic branch tier from above, with second-order branches.

D. Detail of plagiotropic branch from above showing pronounced anisophylly.

A and C from Ektachrome transparencies provided by Dr. J. B. Fisher (Kedah collection); B
and D from Singapore collection.

4

Architecture & phyllotaxis of Anisophyllea disticha

5

Anisophylleae, together with Combretocarpus , Polygonanthus and Poga. These
four genera differ from other members of the family in having alternate, exstipulate
leaves (rather than opposite, stipulate leaves). The other three genera in the tribe
have more limited distribution than Anisophyllea: Combretocarpus (1 sp.) is
restricted to Sumatra and Borneo, Poga (1 sp.) to West Africa and Polygonanthus
(2 spp.) to Amazonia (Pires and Rodrigues, 1971). This tribal aggregation is strongly
supported by evidence from wood anatomy (van Vliet, 1976). A contrasted view is
to segregate a separate family Anisophylleaceae (Schimp.)Ridl. (e.g. Takhtajan,
1969; Airy Shaw, 1973), but van Vliet points out similarities of the Anisophylleae
to other Rhizophoraceae especially in its resemblance to the tribe Gynotrocheae.

It should be useful to establish the comparative phyllotaxis of members of the
family, since the contrast between the tribes in this character is extreme.
Anisophyllea disticha is significant in this respect because of the pronounced
heterophylly of some of its axes which might represent a transition between the
stipulate and exstipulate conditions, especially as the smaller leaves are often
described as “stipule-like”.

Anisophyllea disticha , “leechwood”, is restricted to the Malay Peninsula,
Sumatra and Kalimantan (Borneo) where it is a characteristic understorey treelet up
to 7.5 m high occurring on a diversity of soils, ranging from swampy areas to drier
granitic sands and ridges (Ding Hou, 1958). According to Halle et al (1978, p. 200)
the architecture of some Anisophyllea species represents an extreme expression of
Massart’s model, which refers to trees with an orthotropic, monopodial trunk which
grows rhythmically and produces tiers of plagiotropic branches at regular intervals.
Anisophyllea disticha in particular, stands apart from most other members of the
genus in the extreme plagiotropy of the horizontal branches which are
anisophyllous, with an apparent distichous series of scale leaves superimposed on
the distichous foliage leaves (cf. Tay, 1977). This condition is occasionally found in
A. scortechinii so that there is a possible phyletic link between Anisophyllea disticha
and the rest of the genus (Ding Hou, 1958).

The present study establishes to what extent the phyllotaxis of the plagiotropic,
anisophyllous branches is primary (i.e. determined by their method of inception at
the shoot apex) or secondary (i.e. modified by later re-orientation through differen¬
tial growth or torsions). Previous description has not addressed this problem (e.g.
Ding Hou, 1958) and has not described microscopic details of anatomy of different
leaf types (cf. Geh and Keng, 1974).

Materials and Methods

Two collections of fluid-preserved (FAA) material which included young or¬
thotropic and plagiotropic axes were available for this study (J. B. Fisher, 5.vii.77,
Kedah Peak, Malaysia and P. B. Tomlinson, 5.viii.82, Garden’s Jungle, Singapore
Botanic Gardens).

The distal 1 to 2 cm of a number of both types of axes were embedded in
“Paraplast” and then serially sectioned at 8 p with a rotary microtome. The sections

6

Gard. Bull. Sing. 36(1) (1983)

were mounted on slides and stained in safranin and fast green. In order to facilitate
investigation of leaf development and nodal anatomy the individual sections were
photographed in series through a Wild microscope with a Bolex movie camera, using
the cinematographic drawing method described by Zimmermann and Tomlinson
(1966).

Single transverse and longitudinal sections of mature leaves from these two collec¬
tions were also prepared, using a sliding microtome to cut sections of 20-30 These
sections were mounted in glycerine and examined unstained. Finally, leaves from
these collections were cleared in 5% alcoholic NaOH and examined unstained.

Diversity in external leaf size, form, and disposition was explored by examining
the collections of A. disticha in the Harvard University Herbaria, which consisted
mainly of plagiotropic axes.

Results

Architecture . Massart’s model is well represented by this species in the orthotropic
(trunk) axis which produces a tier of horizontal plagiotropic axes (branches) at wide
regular, intervals (Plate 1A). There are usually 5 branches in each tier (Plate IB).
Our material did not include seedlings but the older trunk axis which we studied
seems similar to the seedling axis described by Geh and Keng (1974). Germination
is described as hypogeal and the plumule bears spirally-arranged scale leaves.
Growth of the axis is rhythmic, eventually with the production of a tier of branches
at the end of each of flush. We have had no opportunity to make extensive
phenological studies but the apex of the orthotropic shoot immediately above the
tier undergoes an extended period of rest before growth is renewed and the next ver¬
tical increment is made. Individuals within a population seem asynchronous with
regard to flushing since orthotropic shoots at various stages of development can be
found at any one time. Field study of marked individuals is needed, however, to
monitor events precisely.

Leaves on orthotropic shoots are spirally-arranged with ^-phyllotactic arrange¬
ment (Fig. 1A, 2A). At maturity the leaves are separated by extended internodes at
the base of the shoot, but they are crowded towards the region of tier insertion.
Leaves are scale-like, usually appressed to the stem and each subtends a minute
lateral bud which normally undergoes no further development (Fig. 2D). Leaves
towards the end of the flush are somewhat larger than those found at the beginning.
Branches (Fig. 1 A, arrows) are produced at the end of each flush by syllepsis (Halle
et al ., 1978), each branch subtended by a scale leaf. Usually five branches are pro¬
duced i.e. one for each orthostichy in the phyllotactic spiral (Plate IB). The subten¬
ding leaves apparently represent the last 5 leaves of the series produced by the apex
of the orthotropic shoot before it undergoes rest. Although there is a short
hypopodium, there is something of a transition in leaf size along the axis at its base
(Plate IB).

Branches extend almost horizontally and are markedly dorsiventral. They branch
infrequently to produce daughter axes of a second and even third order which repeat

Architecture & phyllotaxis of Anisophyllea disticha

1

the plagiotropic organization (Plate 1C). Evidence for rhythmic growth of these
plagiotropic axes is limited since the only articulations are the daughter branches
themselves, which are usually produced in pairs from adjacent internodes (Plate 1 A,
C). No discrete terminal resting buds are produced and there is no regular fluctua¬
tion in size of the two kinds of leaf. Direct measurement of growth frequency is re¬
quired to demonstrate rhythmic growth, if it exists.

Dorsiventrality is determined by phyllotaxis, as shown in serial sections of buds,
with no secondary reorientation of leaves other than their separation by internodal
extension (Fig. 2E-F). Leaves are four-ranked (Fig. 2E) and consist of two ranks of
scale-leaves arranged alternately on the upper side of the branch (dorsal scales) and
two ranks of larger foliage leaves on the lower side of the branch (ventral foliage
leaves). The internodes between them are extended such that in an acropetal direc¬
tion the sequence of leaves is - left scale - left foliage - right scale - right foliage
- left scale ... etc. (Plate ID; Figs. IB & 5A). In terms of the genetic spiral, the

Fig. 1. Anisophyllea disticha (Singapore collection). Morphological details.

A. Apex of orthotropic shoot; sylleptic branches forming terminal tier cut off (arrows).

B. Detail of plagiotropic axis from above to show relative position of branch scales and foliage
leaves; LS-left scale; /?S-right scale; LF-left foliage leaf; /?F-right foliage leaf; branch scale -
stippled (cf. Fig. 4).

C. T.S. trunk axis at level of insertion of trunk scale to show position of multiseriate gland;
vascular tissue-solid black.

D. Detail of multiseriate gland.

8

Gard. Bull. Sing. 36(1) (1983)

following values represent observed angular divergences proceeding acropetally
(n = 11):-

Angle of Divergence

Sequence

Left scale - left foliage
Left foliage - right scale
Right scale - right foliage
Right foliage - left scale

46° ± 13
180° ±0
314° ±13
180° ±0

In Figure 5B, these four sequences are indicated by the numbered sequences 4-5,
5-6, 6-7, and 7-8, respectively.

Scale leaves subtend no axillary buds (Fig. 2K) whereas foliage leaves always sub¬
tend at least one bud (Fig. 21) which usually develops as an inflorescence, but occas-
sionally as a vegetative branch which grows out by syllepsis and, as stated, repeats
the construction of its parent axis.

Reiteration of the architecture in the sense of Halle et al. (1978) is seen in the pro¬
duction of additional orthotropic shoots from the trunk axis, presumably from dor¬
mant lateral buds. This most typically occurs in damaged stems with the resulting
“repair” of the original crown. However, lower buds occasionally grow up without
obvious damage to the plant. Since the architectural crown-form is so precise and
plagiotropic shoots never produce orthotropic axes, the result is always a narrow
cluster of orthotropic axes each with its own set of branch tiers which become
somewhat internested.

Leaf morphology . Three types of leaves can be recognized on the basis of mor¬
phology and position. Symmetrical scale leaves are the leaves of orthotropic shoots,
asymmetrical (dorsal) scale leaves and (ventral) foliage leaves are the two kinds on
plagiotropic shoots (Fig. 3A, C and E). These three types will be referred to as trunk
scales , branch scales , and foliage leaves , respectively. Representative dimensions
from the two fluid-preserved collections are included in Table 1 to show that the two
types of scale leaves are about the same in length. Herbarium specimens confirm
these observations in terms of relative size, but several specimens had unusually
large foliage and scale leaves, up to 80 x 35 mm and 10x5 mm, respectively. Ding
Hou (1958) notes that the size of leaves (referring only to foliage leaves) is
“variable” and comments that “all specimens from the Malay Peninsula have small
leaves”. There may be some overlap in size between leaves of different types, with
large distal scales on orthotropic shoots approaching the size of proximal foliage
leaves on plagiotropic shoots. The two contrasted leaf types on plagiotropic shoots
always retain their relative size difference, with transitional forms restricted to the
branch base.

Mean length of foliage leaves in the two collections varied from 16 mm (Kedah
collection) to 25 mm (Singapore collection) and mean length was in both cases two
to three times greater than mean width. Both types of scale leaves in both collections
were about the same mean length, 5 mm, but trunk scales were fifty percent wider
than branch scales (cf. Fig. 3C and E).

Fig. 2. Anisophyllea disticha (Singapore collection). Phyllotaxis and nodal anatomy.

A-D . Orthotropic shoot.

A. T.S. bud 8 fi below shoot apex.

B-D. T.S. three successive levels to show nodal anatomy.

B. Leaf trace with single leaf gap. C. Branch traces from margin of leaf gap. D. Level of
axillary bud.

E-F. Plagiotropic shoot.

E. T.S. bud at level of shoot apex.

F. T.S. bud 64 /z below shoot apex.

G-I. T.S. three successive levels to show: G. Departure of foliage leaf trace; H. Departure of
bud trace.

J-K. T.S. insertion of branch scale to show: J. Unilacunar node; K. Absence of bud.

Vascular tissue - solid black; branch scale - stippled; axis - cross-hatched.

9

10

Card. Bull. Sing. 36(1) (1983)

Table 1. Comparative leaf dimensions (in mm)

Kedah Foliage 16 ± 2.4 x 7.0 ± 1.3 (n = 42)

Branch scale 4.9 ± 0.6 x 1.1 ± 0.4 (n = 47)

Trunk scale 5.2 ± 1.8 x 1.7 ± 0.5 (n = 7)

Singapore: Foliage 25 ± 2.2 x 9.0 ± 1.1 (n = 33)

Branch scale 5.3 ± 0.4 x 1.7 ± 0.3 (n = 33)

Trunk scale 5.5 ± 1.4 x 2.5 ± 0.9(n=17)

Foliage leaves (Fig. 3A) are rhombic but asymmetrical, with an acute tip and an
acute, unequal base; branch scales (Fig. 3C) are asymmetrical, lanceolate-ovate but
somewhat falcate with an acute tip and an acuminate or rounded but unequal base.
Trunk scales (Fig. 3E) are almost symmetrical, with an acute tip and an attenuate
or rounded, unequal base. The margins of all leaves are entire, those of the foliage
leaves sometimes slightly inrolled abaxially.

The venation of foliage leaves may be described as acrodromous (Fig. 3A) with
a mid-vein which is prominent below (Fig. 3G). Two pronounced secondary (lateral)
veins originate basally from the mid-vein, with a third lateral vein originating
suprabasally from the anodic or acroscopic side of the lamina. All three secondary
veins run almost to the apex and are connected by regularly arranged cross-veins
(tertiaries) which form a scalariform pattern (Fig. 3B). Venation of both types of
scale leaf is much simpler and may be described as camptodromous and
cladodromous (Fig. 3D, F). The single mid-vein gives rise to minor secondaries
which extend towards and sometimes along the margins and interconnect only occa¬
sionally to form an open reticulum.

Both leaves and stems are covered with fine, brown, apparently uniseriate hairs.
These differentiate early, when the leaf primordium is at about the fourth to sixth
plastochron. Each originates from a four-celled basal complex of cells which gives
rise to a single elongated distal cell, which becomes thick-walled in the mature leaf.
The indumentum of the two types of leaf on plagiotropic shoots is somewhat con¬
trasted, since hairs are usually absent from the scale leaves except along the leaf
margins whereas the foliage leaves are quite densely hairy, on the surface but
especially along the margins. Trunk scale leaves have frequent abaxial hairs,
especially at the base of the lamina, as well as marginal hairs.

Multicellular clavate glands, in pairs or sometimes fours occur at the base of each
type of leaf in a stipular position (Fig. 1C, D). They are not vasculated, and develop
precociously so that they are conspicuous in the buds.

Leaf anatomy. Leaves of the three types are of comparable thickness, measured
halfway between base and apex and mid-rib and margin, shown in Table 2, but
mesophyll structure varies appreciably.

G

1 mm

0.1 mm

Fig. 3. Anisophyllea disticha (Singapore collection). Leaf anatomy.

A-B; G-H. Foliage leaf.

C-D; I-J. Branch scale.

E-F; K-L. Trunk scale.

A, C, E. Leaf outline showing major veins, from cleared specimens.

B, D, F. Details of venation in same leaves.

G, I, K. Outline of complete leaf (half-leaf in G).

H, J, L. Detail of mesophyll anatomy.

Colourless hypodermis - lumen dotted in FI and K.

12

Gard. Bull. Sing. 36(1) (1983)

Table 2. Comparative Leaf Thickness (in microns) (n= 10)

Foliage

130 ±

11

Branch scale

140 ±

18

Trunk scale

140 ±

19

All leaves have a thin, smooth cuticle. The epidermal cells have a sinuous outline
in surface view. Stomata occur on all three kinds of leaves and are restricted to the
abaxial surface, except for a few adaxial stomata towards the leaf apex in the trunk
scale leaves. As shown in Table 3, differences in stomatal distributions are greatest
between foliage leaves and branch scales. Stomata are each surrounded by 4-7
epidermal cells with somewhat less sinuous anticlinal walls.

Differences among the three leaf types occur mainly in the mesophyll. Foliage
leaves (Fig. 3G, H) have a single almost continuous, colourless abaxial hypodermal
layer and a well-differentiated mesophyll consisting of a single adaxial palisade layer
and 2-4 layers of spongy mesophyll. Trunk scale leaves (Fig. 3K, L) may also have
a single layer of colourless hypodermal cells below the abaxial surface (Fig. 3L) but
in neither type of scale leaf is the central mesophyll of 2-4 layers differentiated into
palisade and spongy tissue (Fig. 3J, L). Branch scale leaves (Fig. 31, J) are
distinguished by the numerous epidermal tannin cells, staining dark red with
safranin, which differentiate early but become less conspicuous in mature leaves.
Tannin cells are otherwise common in the mesophyll of all three kinds of leaf. The
general conclusion is that the biggest differences in leaf anatomy are between the
two types of leaf on plagiotropic axes.

Table 3. Comparison of stomatal frequency and stomatal index in different
leaf types (n -10 for each leaf type).

Foliage Leaf

Branch Scale

Trunk Scale

Stomatal frequencies

93 ± 10

55 ± 18

70 ± 18

(stomata/mm 2 )

Stomatal index

.060 ± .008

.049 ± .012

.061 ± .013

(stomata/epidermal cells)

Nodal Anatomy . Cine-analysis shows that all leaves have a single leaf trace deriv¬
ed from a single leaf gap (Fig. 2B-D; G-I; J-K). Geh and Keng (1974) report the
nodal anatomy for the family as M many-traced, trilacunar”, and thereby imply that
this also applies to Anisophyllea disticha. This is a curious discrepancy but it is
possible that these authors refer to the two lateral traces from the upper margins
of the leaf gap which are actually traces to the axillary bud and not the leaf (Fig.
2C, D; H, I).

LENGTH IN MM

Architecture & phyllotaxis of Anisophyllea disticha

13

Internode length . Evidence was sought to show that the scale leaves on
plagiotropic shoots had a constant association with a foliage leaf, since this could
relate to their interpretation as stipules. Figure 4 shows that in the Singapore popula¬
tion (lower histogram) there was an average greater length of the internode between
a scale leaf and foliage leaf on the same side of the axis than between scale leaf and
foliage leaf on opposite sides of the axis. However, this relationship was reversed
in the Kedah sample (upper histogram). Herbarium specimens provided further ex¬
amples of variation in spacing between the two types of appendage. In many ex¬
amples, spacing is about equal. This leads to the conclusion that there is no constant
association between the scale leaf and adjacent foliage leaf which might assist in an
interpretation of the distinctive phyllotaxis.

INTERNODE

Fig. 4. Anisophyllea disticha . Histogram of average internode length along plagiotropic branch in two
collections (Kedah, Singapore); vertical lines above bars represent standard error, numbers within
bars represent sample size. Internodes are those between left scale and right foliage leaf ( LS-LF );
left foliage leaf and right scale ( LS-RS)\ right scale and right foliage leaf ( RS-RF ) and right
foliage leaf and left scale {RF-LS)\ cf. Fig. 2B.

14

Card. Bull. Sing. 36(1) (1983)

Discussion

Phyllotaxy. Evidence from nodal anatomy demonstrates that the 3 kinds of leaves
in Anisophyllea are homologous since they all have the same unilacunar, single-trace
vascular configuration, and therefore cannot be regarded as referring to two
categories, v/'z. leaf and stipule, even though accounts of the branch scale leaves
have referred to them descriptively as either “stipule-like” (Ding Hou, 1958) or “ap¬
pearing as stipules” (Corner, 1952). Nor are they consistently associated with foliage
leaves in a way which might suggest that they are leaf-opposed stipules (Fig. 4). Fur¬
ther evidence is provided by the similar anatomy of scales on orthotropic and
plagiotropic shoots (Fig. 3). The homology between trunk scales and leafy appen¬
dages is not disputed since these scales are solitary, with a normal spiral phyllotaxis
and subtend axillary buds. Morphological evidence that the branch scales are
stipules therefore comes solely from the absence of axillary buds. The suggestion
that they might be stipular can otherwise be based only on out-group comparison
since stipules are a characteristic feature of other tribes within the Rhizophoraceae
(e.g. Rhizophoreae, Gynotrocheae). The microscopic glands at the leaf base are not
stipule homologues despite their position (Fig. 1C).

If we rule out the possibility that Anisophyllea possesses stipules, the phyllotaxis
of the different shoot systems becomes easier to interpret. The orthotropic shoots
have a regular spiral phyllotaxis (Fig. 2A) and offer no interpretative problems.
However, the four-ranked leaf arrangement of the plagiotropic shoots (Fig. 2E) is
difficult to interpret as a modification of any more regular or familiar phyllotaxis
without envisaging major rearrangement of appendages. Each of these major ar¬
rangements may be considered in turn: -

1. Spiral. To accommodate a genetic spiral of the type found in spiral phyllotaxis,
in which the angular divergence between successive leaves in the spiral is some
regular Fibonacci fraction of the total stem circumference, would require modifica¬
tion of each leaf postion. A change from 144° (the Fibonacci angle) for a-|-
phyllotaxis would have to result in different displacements for each successive leaf
(cf. the values on p. 8). Such a modified spiral is a difficult interpretation to accept.

2. Distichous. A distichous leaf arrangement in the plagiotropic shoots could only
exist if it is hypothesized that two independent sets, of distichous leaves are present.
Two such arrangements can be hypothesized. Following one interpretation, one
series would consist of foliage leaves (leaves 1-3-5-7 in Fig. 5A, B), and the other
series would consist of scale leaves (leaves 2-4-6-8 in Fig. 5A, B). The orthostichies
of these separate series would be set an angle of approximately 135° to each other
(Fig. 5B). The sequence of initiation of appendages which is continuously acropetal,
without any evidence for the suggested two separate series, does not support this in¬
terpretation. Furthermore, this interpretation is purely descriptive, and does not ex¬
plain the phyllotaxis as a modification of a more standard arrangement.

Alternatively, the two series could each consist of a scale and a foliage leaf on op¬
posite sides of the stem, that is one series of left foliage-right scale (leaves 1,2,5,6
in Fig. 5A, B) and the other of right foliage-left scale (leaves 3,4,7,8 in Fig. 5A, B).
The orthostichies of the two series in this case would be set at an angle of approx-

Fig. 5. Anisophyllea disticha (Singapore collection). Leaf arrangement on plagiotropic shoot
(Diagrammatic).

A. Plagiotropic axis from above, with foliar appendages (both foliage leaves and branch scales)
numbered acropetally from 1 (oldest, basal) to 8 (youngest, distal). Odd-numbered appendages
are foliage leaves; even-numbered appendages are branch scales (stippled).

B. Diagram (not to scale) of axis (cross-hatched) shown in A, with the eight numbered appen¬
dages projected onto a transverse plane, in their four orthostichies. Angles between orthostichies
are approximate (see text).

imately 45° to each other (Fig. 5B). This interpretation requires that the sequence
of leaves of each series be interrupted alternately by 0,2,0,2, etc., leaves of the con¬
trasted series. This is hardly a standard distichous arrangement.

3. Decussate and bijugate. Particularly difficult problems exist if the phyllotaxis
of the plagiotropic shoots is considered to be a modification of a system in which
leaves are opposite with pairs either at right angles (decussate) or some other regular¬
ly repeating angle (bijugate). The last arrangement needs to be considered for com-

15

16

Gard. Bull. Sing. 36(1) (1983)

parative purposes because the phyllotaxis of the Rhizophoreae is of this kind
(Tomlinson and Wheat, 1979) and it seems also to occur in at least some of the non¬
mangrove genera (e.g. Carallia, Crossotylis , Gynotroches , Pellacalyx). In Figure
5A, B, a bijugate arrangement would be represented by pairing a foliage and a scale
leaf in the fashion 1-2, 3-4, 5-6, 7-8. However, to accommodate such a modification
would require internodes to be separated, the bijugate angle to be changed both in
magnitude and direction and the development of anisophylly between members of
a pair.

Adaptive considerations . The conclusion is therefore that there are constant dif¬
ficulties in seeking for a hypothetical ancestral form of shoot with any kind of or¬
thodox phyllotaxis from which Anisophyllea disticha might be derived. The present
conclusion must be that the leaf arrangement on the plagiotropic shoots is unique
and represents a departure from those arrangements usual in angiosperms. A careful
examination of architecture in other species of Anisophyllea might throw light on
the subject.

Contrasted leaf arrangements on plagiotropic and orthotropic shoots are not
unusual in tropical trees conforming to Massart’s and Roux’s models (Halle et al. y
1978) but do not usually take the extreme form found in Anisophyllea disticha . It
is reasonable to interpret them as mechanisms for maximizing photosynthesis by
maximizing the surface area exposed and minimizing overlap both among leaves within
one branch complex, and, between branch complexes of different tiers. The adap¬
tive success is particularly appreciated because Anisophyllea disticha is a treelet of
forest understorey and seemingly able to survive under a closed canopy under
relatively dense shade, At the same time it is, according to Ding Hou (1958), the
most widespread species of Anisophyllea , both geographically and ecologically. Giv-
nish (1979) has hypothesized that optimizing the shape of leaves in planar arrays in
order to minimize overlap will result in a leaf that “must be modified from a wedge-
shaped form to one in which the leaf margin roughly parallels the midrib over much
of the midleaf and tapers toward either end. The apical section should remain
roughly wedge-shaped”. His diagram of such a theoretical system (Fig. 6) shows an
uncanny resemblance to the arrangement and shape of foliage leaves in
Anisophyllea disticha (cf. Fig. IB). Givnish suggests that this modification of leaf
shape from wedge to parallelogram results in a more efficient “ribbon of photosyn¬
thetic tissue along the branch”. Anisophyllea disticha seems not only to have
brought this theoretical model to life, but to have improved upon it as well. Scale
leaves further increase the surface area of the “ ribbon of photosynthetic tissue”
by patching the holes within it i.e. filling the gaps between the bases of the foliage
leaves as well as covering the dorsal side of the branch. This interpretation does not,
of course, rule out the possibility that the scale leaves have some other function not
associated with photosynthesis. For example, they may be involved in close-packing
of appendages within the dorsiventrally flattened shoot apex as a consequence of its
apparently continuous growth.

Although the phyllotaxis of these plagiotropic branches may be unique within the
angiosperms, the same leaf arrangement occurs in other groups. Several writers have
commented upon the considerable similarity between species of Selaginella (section
Heterophyllum) and Anisophyllea disticha (e.g. Corner, 1952). However, in
Selaginella leaves are in pairs. Dengler (1983a, b) has recently described leaf

Fig. 6. Theoretical optimal leaf shape and display along one side of a horizontally orientated branch seen
from above; leaves shown on one side of the branch only (after Givnish, 1979).

development in plagiotropic shoots of Selaginella martensii which shows a similar
dorsiventrality determined by primary leaf arrangement. Her studies were under¬
taken to show that appendages which are “homologous” diverge considerably from
each other very early in ontogeny and that the development of small dorsal leaves
cannot be regarded as a simple truncating of the developmental process which oc¬
curs in larger, ventral leaves.

Our initial observations lead to the same conclusion in Anisophyllea disticha since
dorsal and ventral leaves can be distinguished histologically within three to four
plastochrons of their inception. At their third plastochron primordia which will
develop into foliage leaves have six cell layers (at a point midway between
longitudinal axis and margin) versus only five layers in scale primordia at the same
developmental stage. Foliage primordia also show evidence of an abaxial ridge cor¬
responding to the position of the midvein (Fig. 2E), and stain more densely with
safranin than do scale primordia. Dengler’s work (1983a, b) provides a model for
the further investigation in developmental differences between contrasted leaf types
in Anisophyllea disticha . Such an investigation would complement our initial obser¬
vations and would contribute considerably to our understanding of the mor¬
phological plasticity of the vegetative parts of higher plants in response to limiting
environmental circumstances.

Biblography

Airy Shaw, H. K. (1973). In J. C. Willis, A Dictionary of Flowering Plants and
Ferns. Ed. 8. Cambridge University Press, Cambridge.

Corner, E. J. H. (1952). Wayside Trees of Malaya in two volumes. Vol. 1, Second
Edition. Government Printing Office, Singapore.

17

18

Gard. Bull. Sing. 36(1) (1983)

Dengler, N. G. (1983a). The Developmental Basis of Anisophylly in Selaginella
martensii. I. Initiation and morphology of growth. Amer. J. Bot. 70: 181-192.

-(1983b). The Developmental Basis of Anisophylly in Selaginella martensii.

II. Histogenesis. Amer. J. Bot. 70: 193-206.

Ding Hou. (1958). Rhizophoraceae. Flora Malesiana. Ser. 1, vol. 5, pp. 429-493.
Noordhoff-Kolff N.V., Djakarta.

Geh, S. Y. and Keng, H. (1974). Morphological Studies on Some Inland
Rhizophoraceae. Gdns y Bull. Singapore. 27: 183-220.

Givnish, T. J. (1979). On the Adaptive Significance of Leaf Form. pp. 375-407 in:
O. T. Solbrig, S. Jain, G. B. Johnson and P. H. Raven (eds.) Topics in plant
population biology. Columbia University Press, New York.

Halle, F., Oldeman, R. A. A. and Tomlinson, P. B. (1978). Tropical Trees and
Forests: an architectural analysis. Springer Verlag. Berlin, Heidelberg and New
York.

Pires, J. M. and Rodrigues, W. A. (1971). Notas Sobre os Generos Polygonanthus
e Anisophyllea. Acta Amaz. 1: 7-15.

Sandwith, N. Y. (1952). Contributions to the Flora of Tropical America LV-
Discovery of Anisophyllea in America. Kew Bull.: 303-306.

Takhtajan, A. (1969). Flowering plants , origin and dispersal. Smithsonian Institu¬
tion Press, City of Washington.

Tay, Eng Pin. Architecture and Growth Dynamics in Rhizophoraceae. Unpublished
B. Sc. Honours Thesis, 1977, University of Malaya, Kuala Lumpur, W.
Malaysia.

Tomlinson, P. B. and Wheat, D. W. (1979). Bijugate phyllotaxis in Rhizophoreae
(Rhizophoraceae). Bot. J . Linn . Soc. 78: 317-321.

Van Vliet, G. J. C. M. (1976). Wood anatomy of the Rhizophoraceae. Leiden Bot.
Ser. 3: 20-75.

Zimmermann, M. H. and Tomlinson, P. B. (1966). Analysis of Complex Vascular
Systems in Plants: optical shuttle method. Science. 152: 72-73.